Satellite rainfall estimation over Thailand

Rainfall plays a fundamental role in the earth's climate system. The measurement of rainfall is essential to our understanding of the hydrological cycle. Satellite observations seem to offer the best possibility for estimating rainfall and provide an excellent complement to continuous monitoring of rain events both spatially and temporally to conventional rain gauge networks or meteorological radars. The development of satellite techniques at better spatial and temporal resolution for accurate rainfall estimation, however, still remains a difficult challenge.; The Tropical Rainfall Measuring Mission (TRMM) is the first satellite dedicated to monitoring and understanding tropical rainfall and designed to provide quantitative rainfall information. To improve the data quality, the TRMM algorithms and products are periodically improved and reprocessed. This dissertation is aimed at evaluating TRMM products as well as improving remote sensing techniques for rainfall estimation. For these purposes, the study will be focused specifically on the area of Thailand.; Evaluations are conducted by comparing both versions 5 (V5) and 6 (V6) of TRMM and other satellite merged rainfall (3B42) with in situ measurements by rain gauges over Thailand. The V5 is computed from the Adjusted Geosynchronous Precipitation Index (AGPI) and V6 is computed using the TRMM Multi-satellite Precipitation Analysis (TMPA) algorithm. Unlike V5 3B42, V6 3B42 is merged with gauge measurements at the sub-monthly scale. The V5 (satellite only product) shows large biases with respect to the Thailand rain gauges (TG) data. Comparison of daily TG with V6 3B42 rain estimates shows improved performance over V5 in terms of error statistics. However, statistical analyses of rain/no-rain detection show that the critical success index (CSI) is basically unchanged between V5 and V6, indicating no improvement in term of the skill of rainfall detection.; In the development of satellite rainfall estimation, an alternative technique, Microwave-calibrated Infrared Split-window Technique (MIST), is introduced by combining the use of microwave and infrared information. The approach is based on the AGPI technique. Additional split-window information from infrared channels is used in the training of the algorithm. The MIST is developed and examined during both rainy and dry months. The MIST provides reasonable performance which is comparable to the V6 3B42, however, the CSI is improved. The technique can be applied to areas where microwave estimates (3B40RT) from TRMM and other microwave sensors and the infrared data from geostationary satellites are available.